Originally published on Substack
I want to borrow the language that domestic abuse advocates have fought to get into law and ask, uncomfortably: what happens if we apply it to a sport where the other partner never chose to be there.
In human law, coercive control is defined as a pattern of behaviours intended to isolate, intimidate, regulate and compel compliance, eroding the victim's autonomy and sense of safety over time. It is not "a bad day" or a single outburst; it is the water you are swimming in. Dressage horses swim in that water from the day they are backed until the day they are too lame, too tense, or too shut down to go on.
What coercive control looks like from the horse's side
If you strip away the language of "classical", "contact" and "connection", the daily reality of a dressage horse is simple:
He cannot opt out of being ridden or handled without meeting more pressure, stronger equipment, or social consequences for the human.
His movement, speed, posture and even facial expression are regulated almost every second he is with a rider, mostly through aversive stimuli: legs, spurs, whip, bit.
The rules of the sport explicitly forbid him from being ridden without a metal bar in his mouth, even when national federations elsewhere allow bitless dressage up to small tour without the sky falling in.
He learns very quickly that expressing pain or fear — tail-swishing, head-tossing, breaking rhythm, "spooking" — brings more pressure, not relief.
From the horse's perspective, this is indistinguishable from what human law now calls a "pattern of behaviour intended to threaten, intimidate, harass, isolate, control, coerce, or compel compliance", just with hooves instead of hands. The fact that we add plaited manes, white bandages and a musical freestyle does not alter the contingency structure his nervous system is living in.
An example: the young horse who "won't go forward" in the arena. He is ridden in a flash noseband with the bit fitted high to stop him opening his mouth. The instructor says "kick him", the rider kicks, he jumps forward, the pressure goes away, and everyone praises the "correction". What has the horse learned? Not "seek the contact"; he has learned that the only way to stop the discomfort is to run through it harder. That is the architecture of coercive control.
From resistance to learned helplessness
The scientific term that keeps surfacing when you look honestly at modern dressage is learned helplessness. It describes a psychological state in which an animal exposed to repeated, inescapable aversive events stops trying to avoid them, even when escape becomes possible.
Researchers have warned for years that traditional training methods which are repeatedly unpleasant and offer no clear way for the horse to make them stop will interfere with learning and compromise welfare. You can see the shadows of this in:
The horse who no longer reacts when the rider pulls his head in, despite clear signs of discomfort when coercive Rollkur-style hyperflexion is used.
The horse who maintains a "round" frame while his eyes are fixed, unblinking, ears rotated back, neck flat and immobile: a posture described in the literature as consistent with depressive, shut-down states.
The horse who performs extravagant trot and canter on command but shows no spontaneous play, curiosity or variation when loose, because movement has become about avoiding pressure, not expressing joy.
When horses were ridden in coercively obtained Rollkur, they showed more behaviours associated with stress, conflict and frustration, and higher heart rates compared to normal poll flexion. They were also more reluctant to approach fear-inducing stimuli afterwards — exactly what you would expect when an animal begins to associate work, environment and rider with inescapable discomfort.
The dressage world has managed an extraordinary sleight of hand here. The same flattened affect, low-initiative performance and "obedience" that would ring every alarm bell if we saw them in a child are rebranded as "submission", "throughness", or the horse "seeking the bit".
Rules that encode coercion
Coercive control in dressage is not just about bad riders or ugly moments; it is baked into the rulebook.
The FEI and most national bodies require a bit for competition and often for warm-up, making a metal device in the most sensitive part of the horse's body non-negotiable, not a matter of horsemanship choice.
Attempts to ride bitless are blocked not on welfare grounds but on the circular logic that "it's not in the rules", even as other federations permit bitless dressage to higher levels without catastrophe.
Bridle numbers and noseband requirements turn the horse's head into a noticeboard and restraint point, with little discussion of the cumulative effect on facial sensitivity and emotional state.
Judging language focuses on the absence of overt error — no resistance, no irregularity, no visible conflict — rather than the presence of positive welfare indicators like loose, swinging backs, soft eyes, or spontaneous seeking of the work.
In human law, the absence of bruises does not prove the absence of coercive control. In dressage, the absence of visible rebellion — no bucking, no bolting, no open mouth — is treated as evidence of ethical training. That is exactly backwards. An animal with no way out is often the quietest in the room.
The Ridden Horse Pain Ethogram work should have been a line in the sand. We now know that many "technical faults" our sport writes off as rider error or poor training — tail-swishing, head-tossing, irregular gait, reluctance to go forward — are strongly associated with pain and musculoskeletal compromise. Our response has largely been to train over the top of them, building straightness, strength and collected gaits in bodies that are quietly breaking down. That is not correction; it is containment.
The ethical fork in the road
If dressage is going to survive as anything other than an elaborate choreography of coercive control, it has to face three truths.
First, the horse did not volunteer. He cannot give informed consent. That means the burden on us is higher, not lower, than in any human sporting partnership. "He wouldn't do it if he didn't like it" is not a serious ethical argument when every exit is blocked by equipment, training history and management practices that make non-compliance dangerous or exhausting for him.
Second, partnership is not a feeling the rider has; it is a standard you can see. A genuine partner horse understands the questions because they have been taught without fear. He offers the answers with energy and eagerness, not because he's trying to find the least painful option but because the work has been carefully built to feel good and solvable. He shows a body and face consistent with positive welfare — swinging back, soft eyes, variable ear position, curiosity before and after work — not a tightly managed performance that falls apart the moment the reins are dropped.
Third, we have alternatives. Consent-based, bitless, pressure-minimising training is not fantasy; it is already happening quietly in corners of the horse world, and some federations have demonstrated that allowing bitless dressage up to significant levels does not collapse standards or safety. The question is not "Can it be done?" but "Why are we so invested in a system that refuses to try?"
Equestrian coaching courses that require beginners to never let go of the reins as an insurance condition tells you everything about which body the system is designed to protect. It is not the horse's.
If we took coercive control seriously in dressage, we would have to redesign judging, rewrite tack rules, and re-educate ourselves to treat many of our current "training problems" as what they are: welfare red flags. We would also have to accept that some horses are telling us, clearly and repeatedly, that this form of sport is not for them — and that listening might be the most classical thing we ever do.